Analysis Of WeChat Official Account Reading Volume Based On Data Mining Algorithms

In recent years,with the rapid development of networks and technology,various self-media platforms have emerged.As a social tool in the new network era,WeChat public account has developed rapidly with its practical convenience and timeliness of publishing information.However,as the public numbers of different functions and categories emerge in an endless stream,the attention of WeChat users is becoming more and more scattered,and the WeChat public number fans are getting harder and harder,and the article opening rate is getting lower and lower.Under this circumstance,researching and grasping the factors affecting the reading of WeChat public account is beneficial to the public to improve the reading of published information,so as to fully exert the media characteristics in the increasingly fierce competition.Through the research and analysis of the WeChat public account reading,it can help the public number users to find valuable information,and provide the corresponding development direction and strategy for the public number operator.At present,the research on WeChat public account reading analysis is still in its infancy,so it is of great theoretical significance and practical value to explore the key factors affecting the reading of WeChat public account.This article starts from the perspective of the WeChat public operator and uses R software to analyze 998 data on a public website.Taking the total readings that can be obtained by the public number for each month as the dependent variable,14 characteristic indicators including the public number type and the posting time are selected,and the decision tree regression,,and random forest are used,support vector machine,linear regression six kinds of data mining algorithm construction model.Through the ten-fold cross-validation method,the fitting results of each algorithm model are compared,and an analytical model based on random forest algorithm is established to influence the total reading.Firstly,using random forest algorithm modeling based on all data sets,select the best combination of indicators with good model fitting,including public number type,public number category,name length,avatar color,daily average number of times,single number of articles,original proportion,video proportion,title length,title punctuation index,headline positive sentiment score,information entropy a total of 12 characteristic indicators.The entire data set is then divided into 800 training sample sets and 198 test sample sets.Using the training set data to establish a random forest model,on the one hand,the forecast set data is brought into the model to predict and analyze the reading amount.It is found that the model can give a reasonable explanation for the reading of the public number,which is in line with the actual situation and has practical application value.On the other hand,the model established by the random forestalgorithm ranks the 12 characteristic indicators that are preferred,and obtains the number of single-issue articles,the number of daily papers,the number of public numbers,the information entropy,the proportion of video,and the original.The proportion and title punctuation index are important factors in the operation of the public number,and corresponding strategies and suggestions are proposed for the operation of the public number.Through the research in this paper,it has laid a theoretical foundation for the further construction and optimization of the WeChat public platform,and provided practical operational strategies and directions.